I was trying to think of alternatives to rod ends for use in suspension members that would still allow adjustment, when I had a random thought- how about a rod end with a non threaded portion to react the shear loads, yet a threaded section to allow adjustment?
Such a item could be fabricated out spherical bearing housing welded to the end of a suitable bolt (see pic below). The shear loads would be reacted by the non threaded section acting on the insert, whilst the item would be retained by nuts on the bolt and adjusted by adding shims under the head of the bolt.


Of course you'd have to be very confident of your welding (and tidy!), and the tolerances on the insert/bolt interface would need to be pretty tight, but barring that is this idea completely cuckoo?

Edit: Just seen Theshed's subject which is a similar item, I guess it could be machined out of solid, which would eliminate the welding problem.

[Edited on 23/6/11 by Badger_McLetcher]

I dont see what your gaining here, it looks like rod end to me
Surely rod ends are ok provided you spec them big enough on the threaded portion?
Sorry if im missing the point

Agree, you are talking about a ball joint to me too. Just buy one.

also when you adjust it by making it longer, there is nothing stoping the bolt moving back into place... you will need a nut on both sides with 2 sets of threads...

Whether or not this sort of effort to avoid putting a threaded rod under shear loads depends on how fussy you are. You would not usually want to side load a rod with a series of little v's cut into it. Clearly worries can be addressed by going bigger and heavier but that is not the most elegant solution albeit by far the easiest. Obviously the bigger and heavier approach is perfectly fine for most folks.

There is a good article by a formula student judge on the web where he explains why he thinks loaded threads are a poor idea.

I am not convinced that a welded joint would be a huge improvement. Stop being a whimp and machine a joint from solid - it will only add a year or two to your build.....

Adjustment by shims is fairly easy to design in if you are starting from scratch. My uprights allow for camber adjustment without affecting the steering.

The idea was, as Theshed has said, to avoid putting a threaded section into shear. You could just use larger rod ends, but you've still got the thread section acting as a stress raiser.
Kwik- you'd use shims between the bolt head and insert to adjust the length.
Unfortunately Theshed I don't really have the facilities to machine this kind of thing at the moment.

Cost wise these would probably end up being more expensive, so probably not worth doing for most people. As I said, just a random late night idea

Just as a discussion point, I'm not expert, but I would have thought that putting a threaded section into shear, should be no worse than putting a plain section equal in width to the narrowest point of the thread in shear?
Presumably the threaded v is not going to have the same effect as scoring a piece of glass?

It's rather dependant on the radius of the bottom of the thread (and so the profile of the thread itself). These points act as stress concentrators, meaning a larger section is required to counter this. You've also got to take into account fatigue with items such as this, requiring a larger safety factor.

Edit: These may explain better-

http://www.roymech.co.uk/Useful_Tables/Fatigue/Stress_concentration.html

http://en.wikipedia.org/wiki/Stress_concentration

[Edited on 23/6/11 by Badger_McLetcher]

How small are you intending to go though?

You are looking at a whole load of work, time and money to save what would amount to a maybe a few grams per rod end, possibly 1kg in total. You could probably by some lighter shoes and socks and save just as much weight.

As has been said the solution that has been reached time and time again in motorsport is to use sperical bearings and then gain the adjustment with shims. It also has the advantage (as already mentiond) of allowing you to adjust the camber without affecting the tracking.